Miguel Aguiar

email:    mcpcaguiar at gmail dot com

I am a doctoral student at KTH Royal Institute of Technology supervised by Karl Henrik Johansson, João Sousa (University of Porto) and Amritam Das (TU Eindhoven).

I'm broadly interested in scientific machine learning for control, and in particular the interaction between system-theoretical properties and learning, as well as applications of differentiable surrogate modelling in control systems.

Previously, I was a researcher and software engineer at LSTS in Porto, Portugal, where I worked mainly on trajectory optimisation and control for autonomous underwater vehicles.

News

• January 2026: Our paper Iterative Training of Physics-Informed Neural Networks with Fourier-enhanced Features has been accepted to ICLR 2026!